Resident Evil 7 Benchmarks

chaosophy · 25 Jan 2017 at 20:25

nashathedog said:
In the conclusion he said "you can sometimes literally see textures being filled into objects."

??? How can it be literally? You either see it or you don't.

Rage was bad for that.

It depends how they where using the word. In the correct formal use, or the more modern incorrect usage.

David230 · 25 Jan 2017 at 21:52

I'm getting 60fps solid @ 4k in the full game. 980ti @ 1600/4000

LoadsaMoney · 25 Jan 2017 at 22:35

Ive got it maxed @ 1440p, capped at 80, and it just sits there, on a Fury, and thats on the 16.3 Hotfix drivers from last March

mattyfez · 25 Jan 2017 at 22:56

Hmm, dodgy alt-truth benchmarks by guru3d then?

Have donald trumps tentacles reached that far already?

Brazo · 25 Jan 2017 at 23:20

Don't know about dodgy maybe poorly timed and perhaps ill judged to release 'strange' benchmarks where a 480 is ahead of a 1070. Still it made the AMD fans feel better for a little while

siriq111 · 25 Jan 2017 at 23:41

Pretty easy. If you have an nvidia card, turn off shadow cache. If you have an amd card turn it on. AMD can deliver much better performance with this option. Matching almost high end nvidia cards from mid range.

AthlonXP1800 · 26 Jan 2017 at 05:50

siriq111 said:
Pretty easy. If you have an nvidia card, turn off shadow cache. If you have an amd card turn it on. AMD can deliver much better performance with this option. Matching almost high end nvidia cards from mid range.

Guru3D now updated tests for 2nd time again with yet different results again saw RX480 and 390X no longer pulled ahead of 1070. Look like Guru3D finally tested it properly after 2nd time.

I bought it for £25 and played it with all options maxed out that included shadow cache turned on saw Resident Evil 7 ran very smoothly with ancient 3770K on 1080p at 130-140fps, 70-80fps on 1440p and 35-40fps on 4K.

Nice to see 4C/8T 3770K ran game so much better at 1080p than 10C/16T 5960X Guru3D tested.

Tommy_Here · 26 Jan 2017 at 06:08

Had a quick 5 mins on my laptop with 680m at the start averaging 75 fps but it looks horrible yet all the settings are maxed out :/ I don't think even desktop 680 gets around 75-80fps I have a custom Bios on it however but still I didn't expect it to look as bad as it does and peform as good as it does

Nasher · 26 Jan 2017 at 09:09

Nvidia drivers have been sucking for ages, so not all that surprising.

nashathedog · 26 Jan 2017 at 09:19

Lokken86 said:
Well at least there is one game that Vega will outperform a TitanXP . I think shader cache is meant to boost performance if you have the memory for it. Is that correct?

Something like that, I think it reduces the demand put on the gpu when it needs to use what's cached and ready.

Strife212 said:
What in god's name has happened to the 980ti in the 4k benchmark?

With AMD it's called Fine Wine, With Nvidia it's called Rank Hooch.

nashathedog · 26 Jan 2017 at 09:32

chaosophy said:
It depends how they where using the word. In the correct formal use, or the more modern incorrect usage.

Fair enough

David230 said:
I'm getting 60fps solid @ 4k in the full game. 980ti @ 1600/4000

LoadsaMoney said:
Ive got it maxed @ 1440p, capped at 80, and it just sits there, on a Fury, and thats on the 16.3 Hotfix drivers from last March

Having it mostly indoors has worked in it's favour regarding performance, I'm really tempted to grab it as it looks pretty good and the visuals do too but seeing as it hasn't got 21:9 support, I'll wait and see if they patch that in or I'll get it in a sale in a year or two.

Gregster · 26 Jan 2017 at 10:20

Nasher said:
Nvidia drivers have been sucking for ages, so not all that surprising.

Have they?

VincentHanna · 26 Jan 2017 at 10:32

fs123 · 26 Jan 2017 at 10:41

siriq111 said:
Pretty easy. If you have an nvidia card, turn off shadow cache. If you have an amd card turn it on. AMD can deliver much better performance with this option. Matching almost high end nvidia cards from mid range.

The Guru3D updated results show the 390X getting 92fps but the original test showed 140fps :confused:

. How is that even possible unless they turned off a performance boosting option just so the results can show Nvidia beating the AMD cards. The new results show the Nvidia cards getting the same fps as the old results but AMD getting far worse...WTF.

Original GURU3D test (Shadow Cache = ON).........................................................Updated GURU3D test - 1080P (Shadow Cache = OFF)

If shadow cache or other setting for AMD gives a huge performance boost then that should be shown in the benchmarks. Highly dubious benchmarks to say the least. It's like the old Doom benchmarks running OpenGL only when we all know Vulkan gives AMD the best performance.

EDIT: Apparently they did turn off shadow cache for all cards in the updated results, essentially gimping the AMD cards with 8GB. The original test with the 390X rivaling the 1080 is the correct result for AMD. Nvidia drivers are to blame for their cards current performance and should be improved in another update since the older driver apparently had better performance. Rather poor benchmarking by some sites.

fenchu · 26 Jan 2017 at 11:11

really like the game been running at 4k and havent noticed any slow down yet. However the start of game trundling through forest up to house hurt my eyes , like trying to focus on a blurry image weird.

DragonQ · 26 Jan 2017 at 11:17

This confuses me. Most benchmarks I've seen at 1440p show the Fury ahead of the 390X, but is there a trend of newer games using more VRAM and favouring the 390X? They tend to be the same price so it's a tough decision between the two.

TNA · 26 Jan 2017 at 11:38

DragonQ said:
This confuses me. Most benchmarks I've seen at 1440p show the Fury ahead of the 390X, but is there a trend of newer games using more VRAM and favouring the 390X? They tend to be the same price so it's a tough decision between the two.

Is VRAM not mainly an issue with textures? Just go Very High instead of Ultra on textures and the Fury X with more grunt should be ahead easily I would imagine.

That said if someone was buying today a card that they will be keeping for a few years, I would defiantly recommend going 8GB.

DragonQ · 26 Jan 2017 at 11:47

TNA said:
Is VRAM not mainly an issue with textures? Just go Very High instead of Ultra on textures and the Fury X with more grunt should be ahead easily I would imagine.

That said if someone was buying today a card that they will be keeping for a few years, I would defiantly recommend going 8GB.

Well that's why I'm trying to hold out for Vega.

Gregster · 26 Jan 2017 at 12:32

fs123 said:
The Guru3D updated results show the 390X getting 92fps but the original test showed 140fps . How is that even possible unless they turned off a performance boosting option just so the results can show Nvidia beating the AMD cards. The new results show the Nvidia cards getting the same fps as the old results but AMD getting far worse...WTF.

Original GURU3D test (Shadow Cache = ON).........................................................Updated GURU3D test - 1080P (Shadow Cache = OFF)

If shadow cache or other setting for AMD gives a huge performance boost then that should be shown in the benchmarks. Highly dubious benchmarks to say the least. It's like the old Doom benchmarks running OpenGL only when we all know Vulkan gives AMD the best performance.

EDIT: Apparently they did turn off shadow cache for all cards in the updated results, essentially gimping the AMD cards with 8GB. The original test with the 390X rivaling the 1080 is the correct result for AMD. Nvidia drivers are to blame for the current performance and should be improved in another update since the older driver had better performance. Rather poor benchmarking by some sites.

Tommy_Here · 26 Jan 2017 at 12:43

fenchu said:
really like the game been running at 4k and havent noticed any slow down yet. However the start of game trundling through forest up to house hurt my eyes , like trying to focus on a blurry image weird.

Funny enough same here on my laptop really blurry and horrible when I'm home from work later I'll have to give it a go some more mess with the settings